Field report
Walk into a chemistry lab in 2026 and you will still find people writing in notebooks. Paper notebooks. The kind with corners that curl. On the bench next to them, an Excel sheet someone's grandmother could have built. Beside that, a $400,000 spectrometer that exports its results to a USB stick. The chemical industry is a $5 trillion machine, and large parts of it still run on stationery. This is the room Albert Invent walks into - politely, with a cloud platform.
Scene A Henkel R&D bench in suburban Düsseldorf. The chemist is wearing safety glasses, blue gloves, and the unmistakable expression of a person about to copy a number off a screen into a spreadsheet. Albert Invent's bet is that, soon, she won't have to.
The problem nobody puts on a slide
Software ate the world. It barely nibbled chemistry. The reasons are unglamorous - regulated industries, fragmented data, a workforce trained to trust beakers more than browsers - but the consequence is enormous. New materials still take 5 to 10 years to develop. Most experiments fail. Most failures are not even written down in a form anyone else can read.
Nick Talken noticed this from the inside. He's a chemical engineer who, with Ken Kisner, ran a 3D-printing-resin company called Molecule. They needed scientific software that didn't make them want to throw their laptops into the autoclave. They didn't find any. So they hired data engineers. Then they hired more. Eventually it stopped being a side project and became a different company.
Albert is fundamentally changing the way we have invented for over 145 years.
- Michael Todd, VP Global Innovation (customer testimonial)
The founders' bet
Most software for chemists has been built by software people who once visited a lab. Albert Invent flipped the cast list. Talken (CEO) is a chemical engineer. Ken Kisner (CRO) sold chemistry for a living. Zack Kisner runs product. Neelesh Vaikhary, the CTO, is the engineer in the room - the one who has to make 15 million molecular structures behave like a database. The premise is unsexy and very correct: if you want chemists to use your software, ship it with the assumptions of a chemist, not the aesthetics of a Slack clone.
Investors agreed. In December 2024 Coatue led a $22.5 million Series A, with TCV, Index Ventures, F-Prime, and Homebrew tagging along. The pitch was almost old-fashioned: a real industry, real money, real software, real customers. No pivot, no consumer hook, no token. Index Ventures called it a chance to reimagine the global chemicals market - "built by scientists, for scientists." It is the kind of phrase that would be embarrassing if it weren't, in this case, literally true.
Cap table cameo Coatue rarely shows up in a Series A. When it does, the company tends to be either ridiculous or correct. Albert Invent appears to be the latter.
The product, in plain English
Albert Invent calls its platform an operating system for chemists, which is marketing-speak that happens to be accurate. The core is Albert OS - an electronic lab notebook bolted to a LIMS bolted to project management bolted to a chemical drawing tool. One tab, not seven. On top of that sits Breakthrough, the AI layer the company launched alongside its Series A and, in good marketing tradition, billed as the first AI built specifically for chemists.
Breakthrough predicts molecular properties before a flask is touched. It generates novel formulations. It optimizes experiments the way recommendation engines optimize playlists, except the playlist is "the next polymer that won't crack at minus 40." A separate module, Ask Albert, lets scientists search their company's accumulated experimental history in plain English - so the formula a colleague invented in 2014 and then forgot is, finally, no longer invented in vain. Around all of this sits a regulatory database covering roughly 400,000 chemical substances and an automated safety-data-sheet generator. Glamorous? No. Critical? Yes.
We're building the operating system for chemists - by chemists.
- Nick Talken, Co-Founder & CEO
The proof
Software companies love to claim transformation. Chemists are not, by training, easily transformed. The reason Albert Invent's numbers are interesting is that they were produced by people who count things for a living. One customer cut a formulation cycle from three months to two days. Across the platform's 3,000 scientists, the company reports a 2x acceleration in speed to market. In October 2025, the consumer-health spin-off Kenvue - of Tylenol and Neutrogena fame - became a flagship deployment, an announcement covered by Chemical & Engineering News, which is roughly the chemical industry's New York Times.
Why this matters A 2x acceleration in chemistry is not a 2x acceleration in app usage. It is, often, the difference between launching a product in 2027 and launching it in 2029. Calendars are the only currency materials companies can't print.
The mission, restated without the deck
Albert Invent's stated mission is to digitalize chemistry and materials science so scientists can invent faster, safer, and more sustainably. The unstated one is harder and more honest: end the era in which the world's hardest scientific problems are handed to people whose primary tool for collaboration is e-mail attachments. There is, in this, a small kind of dignity. Chemists are some of the most patient, careful people in the workforce. They deserve software that respects the units they work in.
Sustainability shows up here, too, though Albert Invent is sensible enough not to lead with it. Better formulation tools mean fewer wasted experiments. Fewer wasted experiments mean less wasted solvent, less wasted energy, fewer truckloads of failed prototype to incinerate. The greenest chemistry is the chemistry you didn't have to run.
The most boring sentence in chemistry is also the most important: we tried this last year and it didn't work. Albert wants chemists to actually remember it.
What you can actually do with it
Record
Electronic lab notebook, instrument integration, and project management - so the experiment, the result, and the next step all live in the same record.
Predict
Breakthrough AI scores candidate molecules and formulations before a chemist mixes anything. Trained on 15M structures.
Discover
Ask Albert queries a company's full experimental history in plain English. Old work stops being lost work.
Comply
Automated SDS generation across roughly 400,000 substances and a global regulatory database that's actually current.
Why it matters tomorrow
There is an obvious version of the Albert Invent story and there is a less obvious one. The obvious version is a competent B2B SaaS company in an underserved category, growing nicely, with $22.5 million in the bank and recognizable logos on the marketing page. That story is true. It is also a little bit boring.
The less obvious version is that materials science is one of the last categories where AI could actually move the world's GDP needle. Better batteries. Lighter composites for planes that burn less fuel. Coatings that don't shed microplastics. Pharmaceuticals that don't take a decade. None of that happens without somebody untangling the data layer underneath. Albert Invent has decided to be that somebody. It will not always be glamorous work. It may, eventually, be the kind of work people study in business schools.
The contrarian read The next great AI company might not look like an AI company. It might look like a chemistry-software vendor in Oakland with a logo no one outside the industry recognizes - yet.
Back to the bench
Return to that Düsseldorf lab. The chemist still wears the safety glasses. The spectrometer is still there. The paper notebook, increasingly, is not. The number she used to copy by hand now lands, automatically, in a record her colleague in Shanghai can pull up before lunch. The AI suggests three formulations she hadn't considered. Two of them work. The third becomes a patent. It is a small, undramatic scene. It is also, multiplied by 142 labs and 30 countries, what an industrial revolution actually looks like up close.
Albert Invent will not call it that. Chemists, mercifully, are not given to grand pronouncements. They will simply keep showing up to the bench, doing the work, and quietly inventing the things the rest of us will use without ever knowing where they came from. The new software, at last, will be keeping up.
